Corticosteroids for reducing tuberculosis mortality in persons living with HIV: a systematic review and meta-analysis using reconstructed individual patient data.

Objective: To assess the effect of adjunctive corticosteroids on mortality in persons living with HIV (PLHIV) being treated for tuberculosis (TB).

Design: Systematic review and meta-analysis.

Data sources: PubMed, CENTRAL and EMBASE through 31 December 2023 STUDY SELECTION: Randomised placebo-controlled trials (RCTs) with published Kaplan-Meier survival curves comparing corticosteroids versus placebo in PLHIV receiving TB treatment.

Quality assessment, data extraction and analysis: Three reviewers independently assessed study quality and extracted data. Reconstructed individual patient data were derived from published Kaplan-Meier survival curves, and a one-stage mixed-effects Cox regression model was used to estimate HRs for all-cause mortality.

Results: Four trials involving 873 PLHIV with three forms of TB (618 meningitis, 197 pleural and 58 pericarditis) were included. Over a median follow-up of 19.3 months (IQR, 15.1-30.2), 367 (42%) participants died. At 12 months after randomisation, corticosteroids were associated with a 67% reduction in mortality (HR 0.33, 95% CI 0.26 to 0.41; p<0.0001) compared with placebo. This benefit was maintained during the full follow-up period, with a 17% reduction in mortality (0.83, 0.68-0.99; p=0.0477). In subgroup analyses, a non-significant trend towards benefit was seen for TB meningitis (HR 0.84, 0.67-1.05; p=0.061, two trials), with unclear effect for pleural (HR 0.90, 0.57-1.41; p=0.643, one trial) and TB pericarditis (HR 0.40, 0.15-1.17; p=0.100, one trial).

Conclusion: Adjunctive corticosteroids were associated with reduced mortality among PLHIV treated for TB in this meta-analysis of four RCTs. Further clinical trials are needed to confirm this finding and inform guidelines on the use of adjunctive corticosteroid in this population.

Prospero registration number: CRD42024500865.